Digital assets are more valuable when they are part of a network rather than when they live in isolation; this is true also for research related assets. RO2SHARE delivers a curatorial infrastructure for Research Digital Assets, A.K.A Research objects. We aim to i) facilitate the definition of "ownership" over these assets and ii) make it possible for researchers to establish relations across these assets. We are using semantic web technology, e.g. RDF, in order to expose the resulting datasets; we are using ontologies, e.g. DC, RO, in order to represent the RDAs and the relations amongst them. Ours is an incremental approach, initially the triplets are related to the RDAs, "ownership" is then added to the graph; depending on the type of RDA, e.g. paper, software, figure, dataset, we are adding related triplets from external sources, e.g. Biotea (RDF for PMC), Colil (citations in context) and DBPEDIA. The resulting data from the brief interaction between the end user and the digital assets being managed over our interface is open and it is exposed over our SPARQL endpoint. This project delivers data infrastructure, the GUI is just for demostration purposes; the UX will be part of our future work. We are using SHARE data because it aggregates several sources; we are using the ORCID API as our authoritative ID provider and validator. Our work reuses datamodels and concepts from the Open Archives Initiative Object Reuse and Exchange and the rmap-project.
Before installing make sure that you have included a conf.py file in the docs/ path. There is an example under docs/conf_template.py. You can copy and edit that file in order to install your local RO2SHARE.
You will need to have a SPARQL Endpoint (e.g. Fuseki Triple Store) up and running in order to run this web service.
- Download Apache Fuseki from the Apache Downloads Web site
- Uncompress it
cd apache-jena-fuseki-3.4.0/
./fuseki-serverThe go to http://localhost:3030 and create a new dataset called 'ro2share'.
sudo apt-get update -y
sudo apt-get install -y python-pip python-dev build-essential libxml2-dev libxslt1-dev zlib1g-dev
git clone https://github.com/LinkingDataIO/RO2SHARE.git
cd RO2SHARE
virtualenv .env
source .env/bin/activate
pip install -r requirements.txt
python run.pyReplace the 'localhost' reference in the SPARQL_QUERY_ENDPOINT and SPARQL_UPLOAD_ENDPOINT values in your docs/conf.py with the IP Address of your Docker Host corresponding to the Network Interface docker0.
git clone https://github.com/LinkingDataIO/RO2SHARE.git
cd RO2SHARE
sudo docker build -t username:RO2SHARE .
sudo docker run -p 8080:8080 -v /tmp:/tmp username:RO2SHAREThe client is available at LinkingDataIO/RO2SHARE-client. Clone it and Include your ORCID APP Client ID in src/environments/enviroment.ts and src/environments/enviroment.prod.ts
git clone https://github.com/LinkingDataIO/RO2SHARE-client.git
cd RO2SHARE-client
npm install
npm startgit clone https://github.com/LinkingDataIO/RO2SHARE-client.git
cd RO2SHARE-client
sudo docker build -t username:RO2SHARE-client .
sudo docker run -p 4200:4200 -v /tmp:/tmp username:RO2SHARE-clientOur is an incremental approach that makes use of linked data in order to expand the graph of triplets that are related to a Research Digital Asset (RDA). Depenidng on the type of RDA, we use different sources. For instastance, the Biotea dataset () could easily be enhanced by adding the triplets asserting the relations between the DOI, a git file and a dataset in figshare. Moreover, as Biotea profiles the concepts in the papers by using semantic annotations, the graph could illustrate the aboutness of the tool as well.
The VIVO dataset represents scholars within academic enviroments. It has been built over an ontology that represents profesors, researchers, outcomes, organization, publications, etc. Several universities are part of the VIVO effort, this makes VIVO an authoritative data source because universities are mantaining the information. Although VIVO is not heavy on ORCIDs, some VIVO records do have an ORCID. For our project, ORCID, identifiers in general, is an anchor that helps us to expand related triplets in our graph. User Story: as a researcher I want to bring my publications from my VIVO profile The dataworkflow could beguing by retrieving data from a VIVO instance; this could be done via a SPARQL query. Then, the retrieved publication could be "claimed" and then the graph could be expanded by adding assertions indicating the relation between the claimed object and, for instance, a poster in Figshare.
These include, universities, research institutes, etc.
User story: the research manager at University X wants to retrive a list of publications with their associated datasets for a
given ORCID OR for a given name, e.g. Mike Conlon.
User story: the research manager at University X wants to retrieve the triplets for "paper-dataset-software" for a
given ORCID OR for a given name, e.g. Oscar Corcho.
User story: the editor of a journal wants to ask the author for the triplets "paper-software-dataset" The triplets can easily be created using this tool/approach.
User story: as a researcher I want to retrieve a list of all my triplets "paper-software-dataset" User story: as a researcher I want to retrive triplets, e.g. "papers with dataset and no software"